How to do Regression Analysis for Likert Scale Data? Ordinal Logistic Regression Analysis

How to do Regression Analysis for Likert Scale Data? Ordinal Logistic Regression Analysis:
In this video, I have explained step by step process to conduct Ordinal Regression Analysis with example
You can find the important assumptions and methods related to Ordinal Regression
1. Model Fitting Information
2. Goodness of Fit
3. Pseudo R- Square
4. Parameter Estimates
5.Test of Parallel lines / Proportional Odds
I have explained the Hypothesis tests for each of the above tests and under what conditions they must be accepted or rejected.
I hope you will like the video, please press a like button, share the video and subscribe the channel for more related videos.

[@sreevidyasankarasubramania733]

Thank you so much. The explanation given is simply superb and flawless. I was fixed with what and how to do it. The concepts explained by you is lucid and is incomparable. As a person dealing with chemistry to understand statistics is difficult. I once again thank you for adding confidence in my work

[@haillatesfaye3579]

Thank you very much. I were got many understand from your explanation on ordinal logistic regression

[@swathireddy8636]

This video is excellent. It explained clearly all the concepts of Ordinal Logistic regression. Thanks once again sir of the wonderful work. Please keep uploading more video in Regression Analysis.

[@MyEasyStatistics]

Thank you

[@houstondallas1179]

What app did you use sir

[@nirupamasharma4389]

Sir, what to do if both independent and dependent variables are ordinal scales, can we use this by considering independent variable oncontinuous scale?

[@dukerayleigh5140]

Thank you very much sir. Actually, i learnt Logistic Regression but for Ordinal Regression is out of my syllabus. I've been looking for this video for a month. Thank you once again sir.

[@pulkitjain7267]

What if the pearson and devience values are different i.e. pearson value p0.05?

[@AsmaliaBusran]

where did you get the exponential value btw? Thank you

[@MyEasyStatistics]

Do you want to do data analysis using Amos in a simple and easy way ? Then, this online course, "Mastering Structural Equation Modeling with Amos: From Fundamentals to Advanced Techniques," is just what you need! the course offer a simple and easy learning experience with plenty of practical examples. Visit the website at myeasystatistics.graphy.com to enroll today. Don't miss the chance to learn everything about SEM using SPSS Amos, from the basics to advanced techniques. Start your journey now!

[@niria4106]

I have a response variable called skin yellowness, which I will measure via a scored color chart, whereby 1 is pale yellow and 15 is orange. I'm not sure if this counts as an ordinal variable, because the scale is numerical and is basically a value for pigmentation (making it a numerical variable) or if it is ordinal, because the score suggests some sort of order. Can anyone help?

[@niria4106]

I have a response variable called skin yellowness, which I will measure via a scored color chart, whereby 1 is pale yellow and 15 is orange. I'm not sure if this counts as an ordinal variable, because the scale is numerical and is basically a value for pigmentation (making it a numerical variable) or if it is ordinal, because the score suggests some sort of order. Can anyone help?

[@farahadilaabdullah4437]

Thank you, excellent explanation. Sir, I have one question. There are 8 items for my DV (all measured by using LIkert scale). Do I need to compute mean and recode into a different variable? Or how?

[@MyEasyStatistics]

Yes

[@hikmetabdulhamid]

Hi. When I did the mean of the items, I got decimals. Is that considered an ordinal data or scale afterwards?

[@joapriltaccad2899]

Good day sir, based on the result that the spss provide, *P0.05 Goodness-of-fitting= Accept the NULL (ypu said that it should be more that 0.05 it is okay in your data) *Parameter Estimates= Brand A and D is not significant variable to the dependent variable *P>0.05 Test of Parallel Lines= Accept the null hypothesis My question is, overall what is the result? should you reject or accept the null hypothesis? Thank you

[@peteragyemanbiney1075]

Great vedio. very informative. Thanks

[@jubimothisha]

Sir.Thank you very much . where is exponential values in the tables? I had a doubt

[@chriss5821]

Thank you so much!!!!! This video helped me out a lot. I'm not even taking a statistics course but I need it for a study I am conducting.

[@d.drummer]

Question: can you still use regression if let's say there's only one "brand" being used for the independent variable?

[@hirakalita1952]

an excellent video. very nicely elaborated. Need more videos regarding impact analysis in Likert scale in SPSS

[@decentdentist]

wonderful detailed explanation.. can u pl explain how did u calculate OR/ B estimate?? Thx

[@agatsu3rd]

Dear sir, I have dependent and independent variables both are categorical Likert scale. How to conduct a regression analysis? all variables are latent variables (unobservable).

[@andayosef9823]

Hello , this is a great video.thank you very much. I tried using Ordinal logistic regression on a survey but got bad results. test for normalty was fine. The Indepentent var is the age by ranges (18-25,26-34,35-41,42-50,51+) the depentent var is question in likret scale (1-6) After running OLR the model fit was 0.118 and the Pearson was 0.006. according to what I understand the model does not fit the analisys. could any one help ? I can send the SPSS file . Anda

[@kii1359]

Thank you so much for your video. I was wondering would it be better to use probit regression for likert scale data?

[@meghanaps4597]

It's really a informative vedio. Please explain how to calculate exponential value .

[@gaya3kawshalya93]

exponential constant (e) to the power B (estimate);;; e=2.718 ;;;; calculation for brand C would be as; e^3.757 = 2.718^3.757 = 42.80309

[@aaronayubasamaila8821]

Please why Pearson and not speaman?

[@sujathaelangovan5552]

Excellent video! Very clearly explained. Thank you.

[@muhabepeter3500]

Thank you so much sir....i had this problem of regression analysis and had no one to turn too...but this help me alot please keep uploading more videos

[@lokeshlochunah2156]

Hello sir, how do i remove the frequencies?

[@sumithaluwalia785]

Sir how you calculated the exponential B, how graphs plotted and in table no I have not found exponential B, if exist how can we calculate. Thanks.

[@mukundasharma8856]

very informative thanks alot

[@mohammadarbabpour5212]

Thank you so much. Great Explanation!

[@madguds]

Very clear explanation. Thank you so much.

[@acharyasony]

Thank you for the amazing video sir.. I have a doubt...Sir I have 2 set of data in likert scale. one set shows the current list of amenities in a guest room and the 2ns set shows the upgraded set of amenities in a guestroom. i have satisfaction data in nominal scale ie yes and no. how should i establish the relation between the set of amenities and guest satisfaction and then can i compare the level of satisfaction if it has increased or not

[@ismahanbbeconomie]

Thank you so much, great explanation

[@kossonouprunelle7576]

Thanks for the explanation. Please, i would like to know what should i do when the Pearson and deviance results are significant (under 0.05) in the Goodness-of -fit table? Is it an indicator that the model do not fit the data, and i should not do regression analysis? Also, If the test of parallel lines is significant (under 0.05), can we still use the results of the regression? In this case, should we choose Generalized linear model option on spss to run the regression analysis. Thanks, your reply will help a lot!

[@mosesnyathi418]

Hi, did you get the answer to this question. My Pearson is p

[@sslife001]

Thank you sir. This video is really informative and productive

[@0350332]

I am happy to find your video. exactly what I need. EXCELENT!

[@beneberuteferra1870]

Thank you very much, great job

[@MegaSuperjunior123]

The best explanation for ordinal regression ever! Thank you so much for your explanation! I have gained a lot of knowledge from your video

[@MyEasyStatistics]

Thank you

[@streetrowdies007]

How data view appears for this analysis can someone share

[@sarb3470]

Very nice!

[@dr.kateline]

Totally perfect !!! Thanks

[@gorigaitano6214]

thank you sir for such a wonderful explanation

[@liannelewis4651]

Sir i have 6 independent factors, 3 independent factors such are age groups, gender and income levels which are multiple choice questions and the other 3 independent factors are questions asking how helpful online consumer review are, how important they are and do they influence purchase decision which are likert scale questions And my dependent variable question is likert scale asking after reading consumer reviews online, how likely are you to buy the smartphone? Also my data is non Parametric. Sir can I still go ahead with ordinal regression in the above mentioned case. Awaiting your reply sir. Sir i have my submissions in two days. URGENT PLEASE DO REPLY SIR.

[@MyEasyStatistics]

Yes

[@dr.dineshvellatt1451]

Thanks. Wonderful information

[@theekshyanuwanthini9656]

You are the only one i found that is usefulll

[@ogochukwuonyeso5768]

Thank you for the wonderful explanation.

[@abumonyeigbo5040]

Thank you for uploading.

[@Hikma_by_Scholarx]

what if dependent and independent both are categorical Likert scale

[@amirabankole4426]

Hey did you find out what Tod o , please replyyy

[@amirabankole4426]

*to do

[@MrChuuu]

Show me your data view, awit sa 'yo lods

[@aayushireshamwala854]

This video was really helpful! Thank you 😊

[@priyarana8718]

Good explanation sir

[@kimtsang9682]

Thank you for the helpful video! What program are you using for the analysis?

[@MyEasyStatistics]

SPSS

[@farnazkoubasi5038]

great. Many thanks

[@harshinikrishnamurthy8406]

Thank you so much sir 🙏🏻 i feel blessed to have found your content. You have such in depth explanation along with basics and interpretation. Amazing.

[@vigneshr2367]

Hi Sir.. A great video.. Coming to the satisfaction levels b/w brands, in case my likert scale is coded as 1-Strongly Agree & 5- Strongly disagree, the inference w.r.t. what -ve sign in estimate will have is highly satisfied and +ve sign in estimate will be highly dissatisfied.Is this the right way of how it should be inferred Sir?

[@houstondallas1179]

Sir what app did you use

[@weynua]

This video is a great video than a have seen before thanks so much 👏 and l have one question my result is significant and but the parameter estimates are negative so how can i interpret

[@MyEasyStatistics]

When the parameter estimates in a regression model are negative, it means that there is a negative relationship between the predictor variable(s) and the outcome variable. Even though the relationship is negative, if the result is statistically significant, it indicates that the relationship is unlikely to have occurred by chance alone.

[@madhurimachaudhuri3282]

Excellent demonstration

[@nayabfaizan318]

Really informative

[@tanukhare583]

excellent video. Your teaching skills and explanation are amazing.

[@MyEasyStatistics]

Thank you, please subscribe the channel and share the video

[@nemob1254]

Thank you sir

[@nayabfaizan318]

My variable has 3 groups of AS SOON AS POSSIBLE LATER NEVER will it be considered ordinal

[@fahmseid3314]

good teaching

[@glambysherry]

Dr. Satish Kumar I am getting a lot of help from your videos. God bless you.

[@MyEasyStatistics]

Please subscribe and share the video's, Thank you

[@glambysherry]

@@MyEasyStatistics subscribed

[@HERMINIOJOSECATUD]

Sir can we use regression with different numbers of respondents? Example: Variable 1 respondent is 100 while Variable 2 is 80?

[@Talent-wv1nq]

What is the relationship between ordinal logistic regression, such as this one, and One-Way ANOVA? How can we apply ordinal logistic regression for Repeated-Measure ANOVA? I am highly inspired by your explanation and deep knowledge. Thank you.

[@theunknownpsy8221]

Hi! How do we interpret the likert scale for this? Like for my research is not motivated to highly motivated. how do i know if male for example are moderately motivated? Its urgent huhu thanks for answering! God bless

[@yohanisaactalastas74]

What app did he used for it?

[@theunknownpsy8221]

@@yohanisaactalastas74 SPSS

[@quintoncarroll5803]

What if the pseudo R Square is 1.000?

[@MyEasyStatistics]

If the pseudo R-squared value in a regression model is 1.000, it indicates that the model perfectly predicts the outcome variable using the predictor variables. In other words, the model explains 100% of the variance in the outcome variable, suggesting a perfect fit. However, such a scenario is extremely rare and often indicates a potential issue with the model.

[@whyologist]

Thank you!

[@abebawgetachew7438]

Thank you so much for preparing this video. It is professional and informative.

[@aswathylal8634]

Exemplary Sir... I am doing an analysis using 4 modes of transport where I am considering 8-10 statements per mode considered at 5 levels.. Can I do the analysis in a single go or should it be done separately? How can I arrive at a consolidated result for each mode like how more preferred compared to other modes ?

[@sethulakshmig2999]

Sir, your class is excellent. You explained very clearly. May I ask you one doubt? How to do validation of the ordinal logistic regression model?

[@saritamariaabarzacancino2189]

Thank you very much for the explanation! I have a question, what tests should I do to measure the influence of a set of variables on the occurrence of a phenomenon in the absence and / or combination of them? I'm working on my thesis and I'm trying to find patterns in the interaction of this variables. I would really appreciate your help

[@husseinaman3749]

I am happy to see this video. thank you!!

[@anubratakundu8836]

for both categorical data in dependent & independent, how we can do ordinal regression, please suggest with example

[@tsegayewedajo949]

I am always satisfied with the way you present you video lectures. They are clear, stepwise and understandable. In here presentation why you did put the IV in factor box? Some say they put IV in factor ox if the IV re than two levels how do you react?

[@Manishdna]

Watched the whole video!

[@shivaniabrol1858]

Wonderful explanation of ordinal regression.If we have ordinal( likert scale) independent variable then it is to be entered in factors or covariate? Please tell..

[@nancywaihiga220]

Very well explained.

[@damnpravinrockzz]

Sir. What to do when Goodness of it is less than 0.05. what I have to do. Please do help me

[@chrispinodinga3396]

Excellent!! Thanks for sharing.

[@esivuesuleiman1717]

Please i need a material to help me with a concise methodology for my project using ordinal regression

[@nomanrasheed9644]

Hi. can we use ordinal logit regression when both independent variables and dependent variable are ordinal variables?

[@MyEasyStatistics]

Yes

[@nomanrasheed9644]

@@MyEasyStatistics how can we interpret effect of ordinal predictor (Likert scale) on another dependent ordinal variable (Likert scale)? if you have any video or any easily understandable material please share with me. thanks in anticipating.

[@mylifeisinhishandsamen4167]

I have likert scales as IV... I wonder if I have to treat each level of the likerr scale as categorical...

[@bittubro9720]

I am just a beginner in statistical techniques but I must say never forget the lecture and ordinal logistic regression.

[@c_annl2934]

Sir, how do we write a statement about the result? Please help.

[@jlo6350]

Hello, sir. When no predictors showed significant associations, do we still have to explain the estimates and Exp(B) in our research paper?

[@PradeepKumar-ql5cz]

🎉🎉🎉🎉

[@tamalikasikder5066]

Sir, can you share the data view for the above.

[@skillsmasher4069]

wonderful and amazing job.

[@shashwatasamya9996]

Will it be the same process for two sets of ordinal data?

[@ecsuzinabieta4016]

when test of parallel line less.05, what will next do ?

[@amphitriteymbault3783]

May i ask how do you input the data?

[@pmtwijayasinghe9389]

Sir, This is very useful. I have a question.. I have 5 independent variables, 1 dependent variable. all were measured using lickert scale. My dependent and independent variables are multi dimensional. that means one X variable has 3 dimensions and each dimension has several questions. In that scenario, Can I sumup all the items in all 3 dimensions and arrive at one variable? if so can I use ordinal regression pls?

[@MyEasyStatistics]

You must do Ordinal Regression only when the dependent variable is Ordinal. When you sum the variables then it will become continuous scale data. You can not use Ordinal Regression. You must use Linear regression for Continuous scale dependent variable.

[@emekaoputa4963]

Thanks so helpful

[@user-xt9js1jt6m]

wonderful explanation sir. I have doubt sir how to interpret threshold?

[@mahgolnowparvar6960]

Thank you its a great video ! If there are two or more dependent variable how to identify which variable is the most significant variable among all other variable in survey data ? (likert scale)

[@MyEasyStatistics]

In SPSS we can not do analysis if we have two or more dependent variables

[@shilpasewak3096]

Thank you for explanation, very informative video. Sir, can you please tell, how you derived exponential B values?

[@gaya3kawshalya93]

exponential constant (e) to the power B (estimate);;; e=2.718 ;;;; calculation for brand C would be as; e^3.757 = 2.718^3.757 = 42.80309

[@ivanmuldong7195]

How can I input data

[@NKDuisburg02]

Thank you!

[@irisyvettecesista5966]

Can thia be done using excel?